Conventionally, there is a general object recognition (object recognition) technology in which feature amount of a target object extracted from an image data obtained by photographing the object with an image sensor device is compared with data for collation (feature amount) stored in a prepared dictionary to obtain a similarity degree, and the category and the like of the object is recognized (detected) according to the similarity degree. Moreover, a store system in which such a general object recognition technology is applied to the recognition of a commodity such as vegetables and fruits and the sales of the recognized commodity is registered has been proposed.
In the general object recognition described above, the image data obtained by photographing a commodity varies according to the commodity holding manner, such as a surface to which the commodity is held, a position of the commodity, and a distance between an image sensor and the commodity.
Consequently, in some cases, the commodity cannot correctly be recognized according to a commodity holding manner.